Indonesia’s credit default swaps are pricing a loss of investment-grade status. The rupiah is at IDR 18,190 – a record low. The Jakarta Composite is the world’s worst-performing major equity index in 2026, down 42%. The Hormuz closure contributed to all three. It caused none of them.

Three policy decisions, each taken before the first US-Israeli missile struck Iran on 28 February, dismantled the mechanisms that would have absorbed the shock. What followed was not bad luck meeting a fragile economy. It was a predetermined arithmetic arriving on schedule.

On 12 June, 1,500 students marched on Jakarta’s Hotel Indonesia roundabout under a banner reading “Heading to Bankrupt Indonesia.” Two days earlier, Pertamina had raised Pertamax fuel prices 32%. The students were late to a conclusion the bond market had already reached.

Three Decisions Removed the Buffers Before the Storm Hit

Indonesia’s oil exposure is direct and quantifiable. The country sources approximately 19% of its crude imports through the Strait of Hormuz. Every USD 1 above the 2026 budget assumption of USD 70 per barrel adds IDR 10.3 trillion in subsidy costs. It returns only IDR 3.6 trillion in revenue. At sustained USD 100 oil, the annual shortfall exceeds IDR 300 trillion.

A solvent government with functioning FX mechanisms absorbs that. Indonesia no longer had either.

The first decision was Danantara. Earlier in 2026, President Prabowo Subianto directed commodity exports – Indonesia’s primary automatic foreign exchange stabiliser – into a sovereign wealth fund reporting directly to the presidency. The structure bypassed Bank Indonesia and market-rate pricing.

Kieran Curtis, Head of Emerging Markets Local Debt at Aberdeen in London, said the restructured arrangement was simply “not as efficient as exports finding their own market.” When the rupiah needed dollar inflows, the channel that had historically delivered them was no longer open.

The second was Bank Indonesia’s independence. Parliament passed legislation adding employment and growth objectives to the central bank’s mandate and extending parliamentary powers over monetary policy. Prabowo had also nominated his nephew as deputy governor. Foreign investors had priced BI’s autonomy as the structural anchor for Indonesian fixed income. That anchor was loosening before oil moved.

The third was fiscal headroom. The free meals programme – IDR 268 trillion (USD 15 billion) per year, 14% of the entire 2024 budget – was funded not by new revenues but by cuts to infrastructure and development spending. The absorptive capacity a shock required was gone before the shock arrived.

When Hormuz closed, all three failures activated at once. A falling rupiah raises the IDR cost of dollar-denominated obligations, which widens the fiscal deficit, which drives CDS spreads higher, which triggers capital outflows, which weakens the currency further. The oil price was the trigger. The policy sequencing was the cause.

The Currency Is Not Falling. It Is Feeding on Itself.

The rupiah is down 8% year-to-date and 7% from when the Iran conflict began, with the steepest three-week decline since 2020. Foreign holdings of Indonesian government bonds have collapsed from near 40% before the pandemic to 12.6% – a near 20-year low. Net foreign equity outflows reached USD 3.2 billion through end-May, the heaviest since 2009. Bank Indonesia delivered a 50-basis-point emergency rate hike in May and deployed USD 12 billion in reserves. Neither arrested the slide.

John Woods, Asia Chief Investment Officer at Lombard Odier, was unambiguous: “It is true, there is a doom-loop forming.” Persistent outflows at multi-year lows in bond and equity holdings, he noted, would continue to pressure the rupiah, liquidity and asset prices.

Tan Altundag, Investment Manager for Emerging Equities at Pictet Asset Management – which has reduced its Indonesian equity holdings – was equally direct: “Indonesia is suffering from a genuine confidence crisis.” The currency’s trajectory, he added, risks pushing up inflation, tightening financial conditions and compressing growth in sequence.

A currency under this momentum does not stabilise through central bank signalling. It stabilises when the policies that destroyed confidence are reversed. That has not happened.

A Downgrade Is Not the Same Event for Every Investor

Moody’s and Fitch have cut their Indonesian debt outlooks to negative, citing deteriorating policymaking credibility. S&P has conditioned its rating on fiscal consolidation – a condition the free meals programme and subsidy overhang make structurally difficult to meet. MSCI is reviewing Indonesian equity market standards; a demotion to frontier status remains a tail risk.

The transmission differs by position type.

Government bond holders face the most immediate trigger. A formal downgrade below investment grade activates mandatory liquidation from IG-mandate funds. Foreign ownership stands at 12.6% – a near 20-year low – leaving thin technical support for prices. Forced sellers in a thin market push yields higher, raise borrowing costs and add fiscal pressure at the moment the deficit is already under strain.

The IDX, down 42%, faces a second wave if a credit event triggers EM-mandate exclusions on top of the confidence-driven selling already in progress. The existing decline prices some of that risk. Not all of it.

Private credit and PE positions face a structurally separate problem. Covenant packages written in 2023 and 2024 – when IDR 16,000 was a conservative stress – now operate with the rupiah past IDR 18,000 and no reversal catalyst in view. IDR/USD mismatches that appeared manageable at underwriting are live breaches today. Exit proceeds in USD compress at every point above IDR 16,000, independently of operating performance.

Hemant Mishr, Chief Investment Officer at S CUBE Capital, named the repricing directly: Indonesia is no longer priced as a reliably orthodox emerging market, but as one carrying rising policy risk. The downgrade triggers by position type are mapped in the companion piece: What an Indonesian Downgrade Actually Does to Your Portfolio.

The Exit Requires Undoing What Was Done

The rupiah can stabilise. Indonesia’s resource endowment is intact. A credible reversal on Danantara’s structure, Bank Indonesia’s mandate and the fiscal trajectory would shift the confidence calculus. The conditions are not complex. They are three decisions Prabowo would have to publicly unmake – each central to his mandate, each publicly defended.

Mark Ledger-Evans, Asia-focused Emerging Markets Fixed Income Portfolio Manager at Ninety One, framed it without softening: “It is possible for countries to pull themselves out of a negative spiral where they have put themselves in that position to begin with.”

Possible. Not automatic. The managers pricing stabilisation as remote are not pessimists. They are reading the same political constraints the CDS market priced months ago.

Thousands of students took to the streets in Indonesia’s capital on Friday protesting against government spending plans and fuel price hikes. Demonstrators demanded President Prabowo scrap planned fuel subsidy cuts and redirect military spending toward education and healthcare.

Editor’s View: Every developing country wants fiscal discipline. None wants military cuts – political cost too high. Indonesia’s students are right; government faces impossible choices. The gap between rhetoric and capacity is the story, not the protest.

Read full story →

Three of the world’s largest private banks published global family office surveys between February and May 2026. None of them intended to write the same story. Together, they have.

JPMorgan Private Bank’s 2026 Global Family Office Report surveyed 333 single-family offices across 30 countries, average net worth USD 1.6 billion. Its sharpest finding arrived in four words: AI ambition outpaces allocation. Sixty-five percent plan to prioritise AI. More than half carry zero exposure to growth equity or venture capital – the asset classes where early-stage AI upside is actually captured. Seventy-nine percent hold no allocation to infrastructure, despite its role as the physical backbone of AI through power, data centres and connectivity.

UBS published its Global Family Office Report 2026 on 28 May. It drew on 307 family offices across more than 30 markets, average net worth USD 2.7 billion. Its Southeast Asia finding is the sharpest regional number in the entire survey. Eighty-eight percent of Southeast Asian family offices are already invested in AI. That is the highest of any region globally – 23 percentage points above the 65% world average.

Yves-Alain Sommerhalder, Head of GWM Solutions at UBS, framed the stakes directly: “Artificial intelligence remains the defining investment theme of this decade.” Beyond AI, their top themes are power and resources at 50% and automation and robotics at 44%. Eighty-one percent plan strategic allocation changes within 12 months.

The Conviction Is Not the Problem

Deloitte’s SEA CFO Agenda 2025 surveyed 190 chief financial officers across seven Southeast Asian markets. Seventy-eight percent name AI-related technical skills as their single top concern. Adoption risk, culture and trust, and every other variable on the list follow behind. The capital to invest in AI is present. The engineering capability to deploy it productively inside the operating business is not, at the scale the investment conviction demands.

This tension is not generic. It is structural to Southeast Asia’s ownership model. The dominant corporate architecture across Indonesia, Malaysia, the Philippines and Thailand is the family-controlled conglomerate. Operating businesses and the family wealth vehicle sit under the same patriarch, the same board, the same strategic agenda. When that vehicle commits 88 cents of every AI dollar at the investment level, it is committing to a thesis the group companies must then execute. The two halves of that commitment are running at different speeds.

The WEF’s Future of Jobs Report 2025 put a number on the operating side of the gap. Ninety-six percent of Southeast Asian employers are actively prioritising upskilling – the highest rate globally, against an 85% average elsewhere. They are training because they cannot hire their way out of the shortage.

Demand for AI and machine learning specialists is outrunning supply across every major market in the region. Vietnam, Indonesia, Thailand and the Philippines all cite skills gaps as a leading barrier to AI deployment in the WEF data.

What the Portfolio Reveals

The JPMorgan finding on infrastructure is where the gap becomes a balance sheet risk, not just an operational one. Southeast Asian family offices are heavily committed to AI as an investment theme. Yet 79% of global family offices hold zero allocation to the infrastructure AI runs on – data centres, power generation, connectivity. The same infrastructure gap that constrains AI deployment inside their operating companies is absent from their investment portfolios.

This is not ideologically inconsistent. It is practically significant. A family office with AI software and semiconductor exposure but no infrastructure allocation is betting on the application layer while leaving the foundation unhedged. In Southeast Asia, where grid capacity is the binding constraint on data centre expansion in Vietnam and Malaysia alike, that gap carries direct consequences.

The BNP Paribas and Campden Wealth Asia-Pacific Family Office Report 2025 confirmed that family offices in the region already deploy AI for risk management and investment reporting. But it also identified spreadsheet over-reliance and manual processes as the top internal concerns – overtaking cybersecurity – because the back-office architecture has not kept pace with the investment ambition.

The Execution Window

The CFA Institute’s April 2026 analysis placed Southeast Asia’s intergenerational wealth transfer at USD 5.8 trillion by 2030. The IMF projects 4.3% GDP growth for the region in 2026 – the highest in the world after India. That volume of capital moving between generations within a single decade means the execution gap gets resolved regardless. The question is who closes it first, and on whose terms.

McKinsey and the Singapore Economic Development Board published “AI in Southeast Asia: An Era of Opportunity” in February 2026. The survey covered 330 senior executives across six markets. Nearly half (46%) have moved beyond piloting AI to actively scaling it, placing the region ahead of the global average of 35%. The bottleneck identified is not ambition.

One in five executives named talent as the leading barrier. The specific shortage: MLOps and software deployment skills needed to run AI in production at enterprise scale – not the data science capabilities most organisations have already assembled.

Benjamin Cavalli, Head of Strategic Clients and Global Connectivity at UBS Global Wealth Management, noted that family offices are “maintaining exposure to long-term themes such as artificial intelligence with greater selectivity.” Selectivity means the execution question has arrived.

For managed service providers, AI infrastructure developers and specialist training firms with a regional footprint, the conviction gap is not a risk to manage. It is a commercial opening. The capital is committed. The execution capacity is not built.

The family offices that close that gap inside their own operating companies in the next three years will not be making a bet. They will be collecting the return on one that was already paid for

OECD companies pledged US$ 55 billion for new ASEAN factories in 2022-2023; more than double the US$ 21 billion committed to China in the same period.

The China plus one supply chain strategy that boardrooms spent three years debating had, by 2024, produced a verified outcome: manufacturing FDI into ASEAN hit US$ 44 billion, greenfield investment in electronics and electrical equipment rose 15% to US$ 31 billion, and ASEAN held the top FDI position among developing economies for the fourth consecutive year.

The reallocation is not a diversification exercise. Intel, Samsung, Global Foundries and On Semiconductor have built it into their production architecture. More than 20% of Intel’s global revenues trace back to ASEAN. These are operational dependencies, not portfolio positions.

The ASEAN Investment Report 2025 confirmed the scale, and delivered a judgment the investment promotion materials omitted: “Policy gaps, uneven implementation, skills shortages and infrastructure bottlenecks continue to limit the region’s ability to fully capture supply chain opportunities.”

The gap between capital arrived and infrastructure required to deploy it productively is equally structural. It differs in each market absorbing the bulk of the flow. and it belongs in every underwriting model before the next commitment is made.

Vietnam Manufacturing FDI: The Volume Leader Approaching Its Physical Ceiling

Vietnam posted the region’s strongest manufacturing FDI performance in 2025. Disbursed FDI reached US$ 27.6 billion – the highest in five years – with manufacturing and processing absorbing US$ 18.6 billion, or 59.2% of total registered capital, per Vietnam’s General Statistics Office.

GDP grew 8.02%. The numbers are not accidental. Intel, Samsung, LG and Foxconn all run material operations in the country.

Then Prime Minister Pham Minh Chinh formalised Vietnam’s semiconductor ambitions in Decision No. 1018/QD-TTg, signed September 2024: a roadmap targeting 100 chip design enterprises, a fabrication facility and 10 packaging and testing factories by 2030, with 50,000 engineers to staff them.

The binding constraint is not capital. It is electrons. Power demand for data centres and advanced manufacturing across the six largest ASEAN economies will quadruple from 2.6 GW to 10.7 GW between 2025 and 2035, per Ember Energy analysis cited in the ASEAN Investment Report 2025.

Vietnam’s industrial load is outpacing grid investment. For a PE principal underwriting a Vietnamese manufacturing asset on a five to seven-year hold, grid capacity is not a background risk. It is the underwriting question. The factories are arriving. The power to run them at scale is not.

Malaysia Semiconductor Investment: The High-Value Position with a Talent Pipeline Problem

Malaysia is not competing for Vietnam’s capital. The National Semiconductor Strategy, launched mid-2024, committed MYR 25 billion (US$ 5.3 billion) to front-end fabrication and local vendor development. Malaysia anchors the high-value end of ASEAN’s semiconductor supply chain.

Global Foundries relies on Malaysian facilities for 45% of its capacity at full production. On Semiconductor derives approximately 25% of global revenues from its ASEAN base. These are structural dependencies that place Malaysia’s investment environment in the same risk category as sovereign exposure for the companies carrying them.

The workforce is the constraint. Malaysia’s smaller labour pool and higher wages price it out of volume-oriented manufacturing flowing to Vietnam and Indonesia. The National Semiconductor Strategy targets segments – advanced packaging, front-end fabrication, medical devices – where engineering depth beats headcount.

That is the correct call. The risk is pace. Universities and technical institutions are not producing engineers at the speed the incoming investment requires. Announced capacity and operational capacity are diverging.

For investors pricing Malaysian manufacturing exposure, the talent delivery timeline is where the deal thesis meets its test.

Indonesia FDI: The Scale Argument Carrying an Execution Premium

No other ASEAN market enters the manufacturing reallocation with Indonesia’s combination: 280 million consumers, silica sand reserves anchoring downstream semiconductor input processing, and a national semiconductor roadmap targeting production localisation.

Infineon committed to backend fabrication. Nvidia anchored AI data centre investment. Both bets share the same logic – Indonesia’s resource base and domestic market scale create a manufacturing rationale export competitiveness alone cannot replicate.

The pricing variable is execution. Indonesia ranks consistently among ASEAN’s top FDI announcement destinations and converts those announcements into disbursed, operational capital more slowly than either Vietnam or Malaysia.

The causes are specific: land acquisition timelines, licencing complexity outside major industrial zones, infrastructure gaps beyond Java, and friction between central and regional approval processes.

Kearney’s 2026 FDI Confidence Index, drawing on 507 senior executives, found 84% of global investors rate industrial policy as extremely or very important to investment decisions.

Indonesia’s policy settings are competitive. Its implementation architecture carries a risk premium Vietnam and Malaysia do not – not a reason to exit the allocation, but a reason to price the timeline correctly.

Three Bottlenecks, Three Underwriting Models

Bain’s Asia-Pacific Private Equity Report 2026 recorded APAC PE deal multiples at 13.4 times EBITDA in 2025. At that price, earnings predictability is not a preference. It is a requirement. Vietnam delivers volume and a five-year disbursement track record against a grid ceiling that capital can solve on a defined timeline.

Malaysia delivers value chain position and regulatory consistency. Its ceiling is a talent pipeline that workforce development cannot accelerate on demand. Indonesia delivers scale and resource depth against an execution ceiling that demands a higher IRR hurdle and a longer hold.

Fund managers and PE principals building ASEAN manufacturing exposure are not making one regional call. They are making three separate underwriting decisions – each with a distinct bottleneck, a distinct exit profile and a distinct return threshold. Price them as one thesis and at least two are wrong.

Malaysia’s electricity demand jumped 11.5% year-on-year as rising temperatures and rapid data centre development pushed power consumption to record levels. Gas-fired power generation increased sharply while coal usage declined, with Petronas significantly expanding LNG supply to support growing energy demand.

Editor’s View: Everyone loves announcing data centres. Fewer people ask where the electricity will come from. Malaysia’s power demand is becoming the first real stress test of its AI ambitions, proving once again that digital economies still depend on very physical infrastructure.

Read full story →